868 Journal of Medicinal Chemistry, 2975, Vol. 18, No. 9
Cordes, J . Membr. Biol., 14, 101, 121 (1973); (f) C. Silipo and C. Hansch, Mol. Pharrnacol., 10,954 (1974). (3) (a) J. L. Barker, Nature (London), 252, 52 (1974); (b) E. J. Lien, J. Kuwahara, and R. T. Koda, Drug Intell. Clin, Pharrn., 8, 470 (1974); (c) W. J. Dunn I11 and C. Hansch, Chem,-Biol. Interact., 9, 75 (1974); (d) J. C. Dearden and J. H. Tubby, J. Pharrn. Pharmacol., 26, 73P (1974); (e) S. Inoue, A. Ogino, M. Kise, M. Kitano, S. Tsuchiya, and T. Fujita, Chem. Pharm. Bull, 22, 2064 (1974); ( f ) T. Fujita, K. Kamoshita. T. Nishio-
Riagi et al. ka, and M. Nakajima, Agric. Biol. Chem., 38, 1521 (1974). ( 4 ) W. B. Neely, D. R. Branson, and G. E. Blau, Enuiron. Sci. Technol., 8,1113 (1974). (5) (a) A. Leo, C. Hansch, and D. Elkins, Chern. Reu., 71, 525 (1971); (b) C . Hansch, A. Leo, and D. Nikaitani, J. Org. Chem., 37,3090 (19721. ( 6 ) C. Hansch, A. Vittoria, C. Silipo, and P. E’. C. Jow, J Med. Chem., 18,546 (1975). (7) C. Hansch and S. M. Anderson, J . Org. Chem., 32,2583 (1967).
R , Values of Phenols. Their Relationship with Log P Values and Activity G. L. Biagi,* 0. Gandolfi, M. C. Guerra, A. M. Barbaro, and G. Cantelli-Forti Istituto di Farmacologia e Farrnacognosia, Unioersith di Bologna, Italy. Received December 3 1 , 1974
The experimental R, values for a series of phenols were obtained by a reversed-phase TLC system. The extrapolation from a range of linear relationship between experimental R , values and acetone concentration provided a set of extrapolated R, values. This were used for studying the relationship between structure and activity in vitro and in vivo. The possibility to obtain by means of the extrapolation technique the R , values in a standard system for several series of chemotherapeutic agents is pointed out. The usefulness of the R m values in studying the correlation of chemical structure with biological activity was shown with bis(dich1oroacetamides) and vitamin K analogs,’ penicillins and cephalosporins,2 testosterone esters? and sulfonamide^.^ The chromatographic R, values, as an expression of the lipophilic character of molecules, were shown to be correlated with the Hansch A values in series of testosterone esters3 and sulfonamides.4 In extending the use of a chromatographic technique for the determination of R, values, we turned our attention in this paper to a series of phenols. In particular the purpose of the present work was to show that the present chromatographic technique can provide the R, values of phenols in a standard system, where they can be compared with other series of chemotherapeutic agents. Another aspect of this paper was to further point out the possibility of relationships between partition data in different systems. Finally some experiments were carried out in order to give an additional contribution to the study of structure-activity relationships of phenols. Experimental Section Materials and Methods. The phenols reported in Table I were obtained from commercial sources. The experimental data of the present work were the result of at least 4-8 determinations. R , Values Determination. The chromatographic technique for the determination of the R, values as an expression of the lipophilic character of molecules has previously been d e s ~ r i b e d .The ~,~ polar mobile phase was represented by veronal acetate buffer at pH 7.4 in various mixtures (vlv) with Me2CO. The stationary nonpolar phase consisted of a silica gel G layer impregnated with silicone DC 200 (350 cSt) from Applied Sciences Laboratories. The concentration of Me2CO in the mobile phase ranged from 5 to 55%. The developed plates were dried and sprayed with tetrazotized benzidine.’ After a few minutes at 1 0 5 O yellowish spots appeared on a white background. The R , values were calculated by means of the formula R , = log ( l / R f - 1) Antibacterial Activity Determination. The phenolic compounds were assayed against Staphylococcus aureus 16 R by means of the turbidimetric method. Test tubes containing 4 ml of brain heart (Difco) liquid medium were added with 0.1 ml of EtOH solutions of phenols in order to obtain various serial dilutions. Finally 0.1 ml of an overnight culture of Staph. aureus was added to
each tube. Minimal inhibitory concentrations (MIC) were determined after an 18-hr incubation at 37O. In Table IV the MIC’s are expressed as log 1/C values where C is the average molar concentration which prevents the growth of microorganisms. Hemolytic Activity Determination. The details of the procedure have already been described? A volume of 3.8 ml of phosphate-buffered saline was added to 0.2 ml of a rat erythrocytes suspension. This was obtained by suspending the erythrocytes, separated from 0.8 ml of rat blood, in phosphate-buffered saline to a volume of 8 ml. EtOH solutions of the test compounds were finally added to the system in 1- to 10-rl amounts. Control and test suspensions were incubated for 3 hr at 37O. After incubation all suspensions were centrifuged and the optical densities of supernatants measured at 540 mp in the Bausch and Lomb colorimeter. The results were expressed as percent of total hemolysis provoked by distilled H20 by means of the formula
(ODof sample - OD of EtOH control) OD of distilled t1,O control
X
100
The linear relationship between the concentration of phenolic compounds and the percent of total hemolysis allowed calculation of the molar concentration of each compound provoking a 50% hemolysis. This is expressed in Table IV as log llC. A “t” test showed the statistical significance of the above linear relationship for each compound. LDao Determination in Mice. Albino mice, weighing approximately 20-25 g each, were used. Five animals were injected at each dose level. Dimethyl sulfoxide (MeZSO) solutions of test compounds were administered intraperitoneally. All mice were observed over a period of 24 hr for their death rate. The LDw values were calculated by means of the Spearman and Karber method.8 In Table IV are reported the log 1/C values where C is the LDm expressed as Mhg. Some preliminary experiments were carried out in order to rule out the possibility of toxic effects of MeZSO. Results
R, and Log P Values. The early results of the chromatographic work had shown that the test compounds did not move from the starting line when the mobile phase was represented only by veronal acetate buffer. The addition of acetone was necessary in order to obtain longer migrations and therefore more reliabIe R, values. The plots of the R , values vs. the composition of the mobile phase showed that for each compound there was a range of linear relationship between R, values and acetone concentration. The straight lines of Figure 1, where the data of only a few compounds
Journal of Medicinal Chemistry, 1975, Vol. 18, No. 9 869
R, Values of Phenols
a, ha
3 m *
* 0
m
E 0
m
n
.3 Y
cd
k
c . '
e
m ~ o o o w w e o e o o e o ~ w ~ e o m o ~ m * e o * ~ w o o m w
Q,
0
E 0
0
n
u
1 1 1 9 " " 1 " 4 1 ? 4 1 ?p ?" ?1 p Yo Wo3o?o?o? o o 1d?oYo?d +I P P PooooooPoooo
Q,
E 0
Y
Q)
V
cd .+ v1
m eo
k
cd
* cd
0
eo
v1
scd
d
b
d
m
d
0 4
m
d m N + m 0 + 0 m 0 0 n w m 0
o n m m m o o m o o o o * o o
Y
E
s
* .Y
m
. .
rn E
0
E
a:
2
870 Journal of Medicinal Chemistry, 1975, Vol. 18, No. 9 [ R m + log ( K , + H+)/(H+)] values. In the case of compounds 16-23 the pK, values were not available in the literature and therefore it was not possible to obtain the corrected R m values. Their pK, values were assumed to be close to those of the phenolic compounds with alkyl groups listed in Table I, Le., high enough to make the log ( K , (H+))/(H+)term very small. All this could be supported also by the low u values of compounds 16-23? However, for compound 16 a pK, similar to that for compound 5 would be expected. In this case it would require a significant correction as the log ( K , (H+))/(H+)term for compound 5 is 0.039. In the regression analysis of the present data the uncorrected R m values of compounds 16-23 were used. The log P values in octanol-water were taken or calculated from the data in the l i t e r a t ~ r e . ~ - l ' I t is interesting to note that the range of the extrapolated R m values (from -0.484 to 2.18) is practically as wide as that of the log P values (from 0.80 to 3.64) in the octanolwater system. It would be very interesting to show that, with series of drugs more widely differing in their lipophilic character, the extrapolation technique allows the obtaining of ranges of R,, values comparable to those of the log P values. The relationship between log P and R , values is described by eq l and 2, the latter of which was calculated
20 1.5
18
1.0
A
24
*
10
0
15
e
14
A
13
a
+
+
0.5
0. VI
-u Lo
>
E
a -0.5.
il
-1.0.
log Poct= 1.392 0.993 log Poct= 1.357 1.013
F
s
28 0.960 0.202 (1) 20 0.953 0.192 (2)
1
1
5
(10.145) + (10.111) R , (10.149) + (i0.147) R ,
1.0
1.5
2.0
2.5
3.0
3.5
40
4.5
Acetone c o n c e n t r a t i o n
Figure 1. Linear relationship between uncorrected R , values of phenols and acetone concentration in the mobile phase. The extrapolated R, values of Table I were first calculated from the
equations obtained with the uncorrected data in the range between 15 and 45%and then corrected for ionization. are reported, were calculated by means of regression equations with the R , values in the linearity range between 15 and 45% acetone concentration. A similar linear relationship had been previously obtained for penicillins and cephalosporin^^.^ and testosterone ester^.^ In particular, in the case of penicillins and cephalosporins, since not all the compounds could migrate without the addition of acetone to the mobile phase, the equations of the straight lines were used in order to calculate a theoretical R , value corresponding to a 0% acetone concentration in the mobile phase. As already pointed out, the phenols did not migrate with only the veronal acetate buffer. Therefore, the same procedure described for penicillins and cephalosporins was adopted with the present data. The R , values extrapolated to the system silicone oil-Verona1 acetate buffer are reported in Table I. On the other hand, in a previous work it was shown that the chromatography system silicone oilveronal acetate buffer, i.e., without any addition of acetone to the mobile phase, could provide the R , values of a series of sulfonamide^.^ In conclusion, either directly or by extrapolation the R , values of penicillins, cephalosporins, sulfonamides, and phenols have been so far determined in the same chromatographic system and therefore they could be compared. With acids, such as sulfonamides and phenols, the R, values, in order to be an expression of the true partition coefficient, have t o be corrected according to their degree of ionization a t the pH 7.4 of the mobile phase. In Table I are reported the data necessary for the calculation of the
without taking into consideration compounds 16-23. Equations 1 and 2 are very similar and this should support the assumption made for the pK, values of compounds 16-23. The slopes of eq 1 and 2 are close to unity. According to H a n ~ c h ~this ~ Jcould ~ mean that the stationary phase represented by 5% silicone oil contains the same amount of water as octanol. In a previous work in correlating the R , values of sulfonamides with the .rr values determined in an isobutyl alcohol-water system somewhat lower slopes were found. Again this could be in agreement with the Hansch conclusion that with solvents containing much more water than octanol (i.e,, isobutyl alcohol) there is a slope of less than unity.'*J3 Similarly, a slope of less than unity can be found in correlating the R , values of penicillins, as previously obtained in this laboratory,2 and their log P values determined by S ~ h o l t a n 'in ~ an isobutyl alcohol-aqueous buffer system (Table 11). il
Y
5
log P i - B u= - 0 . 3 3 6 + 7 0.959 0.130 (3) 0.808 R , It should be pointed out that in calculating the regression equations of the straight lines of Figure 1 the experimental data a t 5 and 10%acetone in the mobile phase were not used. This was due to the fact that at these concentrations several compounds showed very short migrations and therefore R , values out of the range of maximum accuracy. Moreover, by considering only the data between 15 and 45% it was possible to obtain straight lines not very far from parallelism. In fact, the extrapolation from higher acetone concentrations should be considered with confidence only in the case of parallelism among straight lines. In order to establish the advantages, if any, of the extrapolation technique the experimental R , values, obtained with increasing acetone concentrations in the mobile phase and corrected for ionization a t pH 7.4, were used for calculating eq 4-12,The extrapolation to 0% acetone in the
Journal ofMedicina1 Chemistry, 1975, Vol. 18, No. 9 871
R, Values of Phenols
Rm
pI-Bu
Dicloxacillin Cloxacillin Oxacillin Phenethicillin Benzylpenicillin Phenoxymethylpenicillin Ampicillin
1.057 0.752 0.588 0.455 0.196 0.130 -0,229
1.63 1.34 1.05 1.03 0.55 0.89 0.08
mobile phase seems to provide R m values better correlated with log P values. The decreasing of the correlation coefficient from eq 4 to eq 12 is due to the fact that at higher acetone concentration all the compounds tend to migrate with the solvent front and therefore have closer R , values. Moreover, higher acetone concentrations will alter the water structure and therefore the character of the partitioning process. Obviously this could decrease the correctness of the extrapolated R , values. The correlation coefficient of eq 1 and 2 could support the usefulness of the extrapolation technique. However, the major point would be in the fair parallelism of the straight lines in Figure 1. The present extrapolated R , values can also be compared with other chromatographic data. Marcinkiewicz et al.15J6 had determined the R , values of many phenols in a chromatographic system where the stationary phase was represented by ethyl oleate and the mobile phase by 25% aqueous ethanol (Table 111). The correlation coefficient provided by eq 13 shows that y
R , = 1.332 (lt0.087) + e0O4
(*0.150) R m ( E t oleate)
2
Y
S
16 0.961 0.141 (13)
there is a highly significant correlation between the R , values obtained in the present work and those determined in the ethyl oleate-25% aqueous ethanol system. In particular, the b's and intercepts of eq 1 and 13 are very close. It can be pointed out that the present chromatographic system allows the obtaining of separate R , values even in the case of halogen-substituted phenols in ortho and para positions. On the other hand, the relationship between I values and AR,,j values of phenols was also shown by Iwasa et al.,17 who took advantage of the R , values obtained in a chromatographic system based on Trigol and diisopropyl ether.13 By means of the data of Tables I and 11, eq 14 was calculated.
R , = 0.070 (10.433) 1.239 (*0.813) R ,
n r S 6 0.902 0.225 (14) prigo1)
r
S
0.940 0.930 0.941 0.931 0.925 0.888 0.870 0.807 0.766
0.192 0.250 0.246 0.264 0.277 0.336 0.325 0.430 0.475
Table 111. Rm Values of Phenols in Different ChrOmatographic Systems
Table 11. Log P and R, Values of Penicillins Compounds
n 28 28 28 28 28 28 28 28 28
~
~~~~
Phenols
Ethyl oleate25% aq EtOH
H
-1.063 -0.471 -0.471 -0.070 -0.070 -0.376
4 -I
0.204 -0.250 -0.444 0.556 -0.906 -0.906 -0.547 0.087 0.395 -1.070
4 -c1 2 -c1 2 -Br 4 -Br 4-CzH5 4 -NO2 2!4 -Mez
3.5-Me2 4-Ph 2 -F 4 -F 2 -Me 4 -17 -Pr 4 -I -Bu 4-OCH3 4-C,H5
Trigoldiisopropyl ether -0.420 -0.585
-0.041
-0.747
0.000 -0.796
Even if calculated with only six compounds eq 14 shows a similar relationship between the present R , values and those of Marcinkiewicz et al.lfi from the Trigol-diisopropyl ether system. The negative slope of eq 14 is due to the nature of the phases in the Trigol-diisopropyl ether system, where the mobile phase seems to be less hydrophilic than the stationary one. Therefore the most lipophilic compounds tend to migrate with the mobile phase and have lower R,, values. In fact, 4-methoxyphenol, which is the most lipophilic compound in the Trigol-diisopropyl ether system, shows the lowest R,, value in the ethyl oleate-25% aqueous ethanol system. Structure-Activity Relationships. The structure-activity relationships regarding the antibacterial activity in Staph. aureus, the acute toxicity in mice, and the hemolytic activity are described by eq 15-20. In calculating these equations the extrapolated R , values of Table I were used. However, because of the good correlation coefficients of eq 4-7 one could have used any of the sets of experimental R,, values obtained with 5-20% acetone concentration in the mobile phase. The introduction of the R,? term into eq 15, 17, and 19 does not improve the correlation coefficient in a significant way. The log l/C values calculated from eq 15, 17, and 19 are reported in Table IV. The deviation of about 2s for com-
872 Journal of Medicinal Chemistry, 1975, Vol. 18, No. 9
Biagr et ul.
Table IV. Data on the Biological Activity of Phenolsa Hemolytic act.
__
Substituent
1
H
2 3 4 5
4 -C1 2 -c1 2.4-C1? 2 -Br 2-CzH5 4-CzH5 4 -NO2 4 -I 2,4 -Me2 3.5-Me, 4-Ph 2 -F
6 7 8 9 10 11
12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
4 -F
2 -hIe 2 -Br.4 -Me 4 - 1 ) -Pr 4 -/ -Bu 2 -/ -Bu 2 - \ -Bu 4 -.\ -Bu 2 -I -Bu. 4 -Me 4 -I -Bu.2 -Me 4 -Br 2 -C1,4-NO, 3 -OH 4-OCHJ 2 -NO,
__________~
----
Log 1 c calcd
2.407 2.588 2.738 3.028 2.689 2,851 2.946 2.600 3.149 2.824 2.894 3.100 2.320 2.556 2.638 2.983 3.228 3.282 3.262 3.374 3.354 3.057 3.307 2.790 3.338 2.100 2.267 2.568
2.439 2.829 2.900 3.057 2.849 2.844 2.837 2.442 3.048 2.805 2.802 3.175 2.481 2.560 2.655 2.763 3.077 3.152 3.286 3.253 3.302 3.339 3.372 2.910 3.104 2.099 2.361 2.483
Log 1/c calcd
1.236 2.523 2.221 3.000 2.523 2.523 2.523 1.921 3.000 2.699 2.699 3.301 1.886 2.221 2.000 2.699 2.824 3.097 3.398 3.301 3.398 3.699 3.523 2.824
1.650 2.487 2.640 2.977 2.529 2.520 2.504 1.655 2.957 2.435 2.428 3.229 1.738 1.910 2.112 2.345 3.020 3.181 3.468 3.398 3.503 3.581 3.653 2.661
0.592
0.917
1.618
_____ ____
1.757
-
-
obsd
obsd
Log 1 / c No.
Acute toxicity in mice Log 1
Antibacterial act. --
-
-
Log 1
.
-
obsd
Log 1 ( talcd
1.744 2.699 2.699 3.301 2.522 2.824 2.921 2.319 3.155 2.522 2.508 3.046 2.097 2.097 2"U9'I 2.678 3.046 3.187 3.398 3.398 3.522 3.522 3.398 2.824
2.046 2.672 2.786 3,038 2.704 2.697 2.685 2.050 3,024 2.634 2.628 3.227 2.112 2.240 2.392 2.566 3.071 3.191 3.406 3.353 3.432 3.491 3.545 2.803
1.521
1.499
2.398
2.126
(
aThe hemolytic and antibacterial activity of compounds 25 and 27 was not determined. Hemolysis of rat erythrocytes 17
log 1/C = 1.414 (10.146) + 1.027 (kO.111)R , log 1/C = 1.329 Lt0.086)
+
?'
S
26 0.959 0.185 (15) 26 0.965 0.197
1.323 (11.326) R , 0.144 (10.148) R; Antibacterial activity in Staph, a w e u s IZ
log 1/C = 1.870 (10.058)
+
0.768 (10.090)R, log 1/C = 1.873 (k0.263)
+
(16) Y
S
26 0.956 0.150 (17) 26 0.955 0.170
0.763 (k0.283)R, i0.002 (k0.128)Rm2 Acute toxicity in mice
(18) It
log 1/C = 2.330 (10.083)
0.478 log 1/C = 2.318 0.524 0.023
+
(10.063)R , (k0.190) + (10.213)R, (10.098)Rm2
Y
S
28 0.928 0.112 (19) 28 0.929 0.134 (20)
pound 16 in eq 15 could be due to the fact that its R , value was not corrected for ionization. Hansch et a1.18 for 15 equations correlating the hemolytic activity of a variety of compounds found a mean slope of 0.93 f 0.17, which is quite close to that of eq 15 (1.03 f
0.11).In particular the slope of essentially unity in eq 15 shows .that the partitioning in the present chromatographic system is closely related to the absorption of phenols by the membrane of the red blood cells, i.e., partitioning between liquid medium and membranes. The best rationalization of the antibacterial activity in Staph. aureus is provided by eq 17 which is characterized by a lower value of the slope. In fact, its slope (0.77 f 0.09) is quite close to that reported for the antibacterial activity of phenols in M . tuberculosis (0.78 f 0.06)12or more generally for the toxicity of various antibacterial compounds to gram-positive bacteria (0.73 f 0.17).18 The interaction with microorganisms seems to depend on R , values in a different way. The reason might be in the lower lipid content of a gram-positive organism such as Staph. aureus. The partitioning in a different chromatographic system could be a better model of the interaction with microorganisms. However, here the importance of the lipophilic character for the antibacterial activity of phenols is confirmed. This is in agreement with Kaye et al.19 who have suggested that the antibacterial action of phenols involves disorganization of phospholipid molecules present in bacterial cytoplasmatic membranes. The relationship between R , values of phenols and their toxicity to mice is described by eq 19.The slope of 0.48 f 0.06 is close to that of 0.4 f 0.1 reported by Hansch12 as typical of systems less sensitive to drug perturbation. In more complex systems, such as a whole animal, the penetration of drugs to their sites of action results from their movement through various membranes and liquid phases. One of the major factors in regulating this random walk
R, Values of Steroids could be the hydrophobic binding to serum constituents. Hansch12 found similar slopes in correlating the log P values of penicillin derivatives with their in vivo activity against Staph. aureus in mice (-0.46 f 0.10) and when relating log P values and hypnotic activity in mice of alkylarylureas (0.55 f 0.09) and 5,5-alkylbarbiturates (0.57 f 0.21). Lien et a1.,20 when studying the acute lethal toxicity in mice of N-substituted lactams and 0-phenyleneures, also found a rather low slope (0.312). Conclusion The present data seem to confirm the usefulness of the
R, values as an expression of the lipophilic character of molecules. In particluar the extrapolation procedure could provide further interesting contributions. In this way the R, values of several series of chemotherapeutic agents could be compared in the same chromatographic system. The highly significant correlation between R, and log P values further points out the possibility of relationships between partition data in different systems. The slope of 0.993 in eq 1indicates the existence of very similar lipophilic characteristics in both systems. Obviously the correctness of the extrapolated R, values seems to depend on the parallelism of the straight lines extrapolating from the linear range of experimental data. In particular it could be important to point out that the extrapolation technique so far provided R, values for several series of chemotherapeutic agents in the same standard system. Moreover, when the A or log P values were available equally good correlations were found with the extrapolated R, values. Finally the results of the structure-activity relationships seem to support the possibility of a classification of drugs according to their mechanism of action a t the physicochemical level. The assay of the same series of drugs in different biological systems, in vitro and in vivo, should be carried out intensively in order to get new achievements in this field.
Journal of Medicinal Chemistry, 1975, Vol. 18, No. 9 873 Acknowledgment. We are grateful to Dr. C. Hansch and Dr. A. Leo for their helpful suggestions. References a n d Notes (1) J. 1).Turnbull, G. L. Biagi, A. J. Merola, and D. G. Cornwell, Biochem. Pharrnacol., 20,1383 (1971). (2) G. L. Biagi, M. C. Guerra, A. M. Barbaro, and M. F. Gamba, J. Med. Chem., 13,511 (1970). (3) G. L. Biagi, M. C. Guerra, and A. M. Barbaro, J. Med. Chem., 13,944 (1970).
(4) G. L. Biagi, A. M. Barbaro, M. C. Guerra, G. Cantelli-Forti, and M. E. Fracasso, J . Med. Chem., 17,28 (1974). (5) G. L. Biagi, A. M. Barbaro, M. C. Guerra, and M. F. Gamba, J. Chromatogr., 41,371 (1969). (6) G. L. Biagi, A. M. Barbaro, M. F. Gamba, and M. C. Guerra, J. Chromatogr., 44,195 (1969). (7) K. Randerath, “Thin-Layer Chromatography”, Academic Press, New York, N,Y., 1966, p 209. (8) D. J. Finney, “Statistical Methods in Biological Assay”, C. Griffin Co. Ltd., London, 1952, p 524. (9) C. Hansch and E. W. Deutsch, Biochim. Biophys. Acta, 126, 117 (1966). (10) C. Hansch, K. Kiehs, and G. L. Lawrence, J . Am. Chem. SOC., 87,5770 (1965). (11) E. J. Lien, C. Hansch, and S. M. Anderson, J . Med. Chem., 11,430 (1968). (12) C. Hansch, “Structure-Activity Relationships”, Vol. 1, C. J. Cavallito, Ed., Pergamon Press, New York, N.Y., 1973. (13) A. Leo, C. Hansch, and D. Elkins, Chem. Rev., 71,526 (1971). (14) W. Scholtan, Artneim.-Forsch., 18,505 (1968). (15) S. Marcinkiewicz, J. Green, and D. McHale, J. Chromatogr., 10.42 (1963). (16) S. Marcinkiewicz and J. Green, J. Chromatogr., 10, 372 (1963). (17) I. Iwasa, T. Fujita, and C. Hansch, J. Med. Chem., 8, 150 (1965). (18) C. Hansch and W. R. Glave, Mol. Pharmacol., 7,337 (1971). (19) R. C. Kaye and S. G. Prandfoot, J. Pharm. Pharrnacol., Suppl., 23,2238 (1971). (20) E. J. Lien, G. L. Tong, J. T. Chon, and L. L. Lien, J. Pharm. Sci., 62,246 (1973).
R, Values of Steroids as an Expression of Their Lipophilic Character in Structure-Activity Studies G. L. Biagi,* A. M. Barbaro, 0. Gandolfi, M. C. Guerra, and G. Cantelli-Forti Institoto di Farmacologia e Farmacognosia, Universitb di Bologna, Italy. Received December 31,1974 The chromatographic R, values of three series of steroids were determined by means of a reversed-phase system. The R, values at 45% acetone in the mobile phase were shown to be correlated with the partition coefficients in an ether-water system. However, an almost equally good correlation was found when using extrapolated R, values. The extrapolation technique could provide a standard system. The relationship between biological data and R, values pointed out the important role of the lipophilic character in regulating the activity of steroids. In particular, the dependence of protein binding absorption and biotransformation on lipophilic character might strongly influence the availability of steroids at the site of action. The R, values of testosterone derivatives and corticosteroids had been shown to be related to the Hansch A values and therefore useful in structure-activity studies.14 In series of phenol^,^ penicillins and ~ephalosporins,6~~ and ster o i d ~the ~ linear relationship between R, values and acetone concentration in the mobile phase has provided the R, values extrapolated to 0% acetone in the mobile phase. This was considered to be a standard system, where all the compounds could be compared. The purpose of the present
work was to study a larger number of steroids in order to compare their lipophilic character. The possibility of the extrapolation in order to calculate the R, values a t 096 acetone in the mobile phase was taken into consideration. The partition coefficients obtained by Flynns in an ether-water system were compared with the present R, values. Finally, some data on the interaction of steroids with the erythrocyte membrane and proteins could further point out the usefulness of R , values in structure-activity studies.
